Spread of varicella in day care is controlled by excluding children at the first signs of illness. Exclusion is generally ineffective. Minimally ill children might be permitted to attend day care. This approach may lead to exposure of those in high-risk groups, i.e., adults or immunocompromised children. Morbidity is greater in adults, but susceptibility among day care workers is probably low. Immunocompromised children can be vaccinated or given varicella-zoster immune globulin (VZIG) after exposure. Questions about the risks of varicella-zoster vaccination (V-Z) concern the production of latent infection and subsequent zoster and the effect on the epidemiology of infection, i.e., its possible delay until adulthood. Zoster has been found not to be more frequent in immunized than in nonimmunized leukemic children, and normal vaccinees retain good antibody titers five years after vaccination and have titers similar to individuals who have had varicella 10 years after vaccination. Vaccine efficacy is excellent, but its desirability will be determined after resolution of questions concerning the long-term impact of its use.